Everyone's portfolio looks great. Every agency says they're senior, agile and communicative. So the real skill in choosing isn't spotting who's good — it's spotting who's good for your project, and who will still be answering the phone six months after launch. We've been on both sides of this: hired for rescue jobs where a previous team disappeared, and turned down projects we weren't the right fit for. This is what we'd tell a friend.

Start with the questions that can't be faked

A polished sales deck tells you how an agency wants to be seen. These questions tell you how they actually work:

"Who exactly will write my code, and can I meet them?"

Some agencies win the work with senior people, then hand it to juniors you never met. Ask to speak to the actual developers. A team that's proud of its people will say yes without flinching.

"What happens after launch?"

Software isn't finished at launch — it's barely started. Ask what support looks like in month two, who fixes a bug at 9pm before your big day, and what it costs. Vague answers here are the single best predictor of pain later.

"Show me something you built that's still running and let me use it."

Screenshots prove nothing. A live product you can click around proves it works, ships and survives contact with real users. If everything in the portfolio is "under NDA" or mysteriously offline, be curious about why.

"Who owns the code and the accounts?"

You should own your source code, your hosting, your domains — everything — outright. Some agencies keep you locked in by holding the keys. Get the answer in writing before you start.

The reference that matters most.

Don't ask for a happy client — ask to speak to one whose project went wrong at some point. How a team behaves when something breaks tells you far more than a glowing testimonial from a project where nothing did.

Red flags worth walking away over

None of these is automatically disqualifying on its own, but two or three together should give you serious pause:

  • A quote with no breakdown. "£30,000 for the app" tells you nothing. A team that understands your project can show you where the hours go. One that can't, hasn't thought about it yet.
  • Pressure to sign this week. Good agencies have a pipeline and don't need to rush you. Artificial urgency is a sales tactic, not a scheduling reality.
  • They agree with everything. If you describe your idea and hear only "yes, easy, no problem", you're talking to a supplier, not a partner. You want the team that pushes back and asks "why?".
  • No mention of testing. If nobody brings up how the thing gets tested before it reaches your users, assume it doesn't.

Local, offshore, or somewhere in between?

UK founders usually weigh a local agency against a much cheaper offshore team, and the honest answer is: it depends on how much hand-holding the work needs. Pure execution of a crystal-clear spec can go offshore cheaply. Anything involving ambiguity, product decisions or tight communication tends to cost more offshore than it looks, once you factor in the back-and-forth, the timezone lag and the rework.

There's a third option people forget: a distributed senior team that combines UK-hours client contact with specialists wherever they happen to be best. That's the model we run — one point of contact in your timezone, real seniority behind it — precisely because it sidesteps the usual local-versus-offshore trade-off.

How to compare quotes without getting fooled

The lowest number wins depressingly often, and it's usually the wrong call. As we cover in our guide to what a web app actually costs, a cheaper quote is frequently just a smaller quote — the same brief with the testing, admin panel and post-launch support quietly removed. To compare fairly:

  • Give every agency the identical written brief, so they're pricing the same thing.
  • Ask each to list what's included and, crucially, what isn't. The exclusions are where the surprises hide.
  • Compare total cost of ownership, not launch cost — hosting, support and the price of changes all count.

Fit matters more than you think

Two agencies can be equally skilled and one will still be right for you. It comes down to how they communicate, whether they treat your budget like their own, and whether they're honest when the answer is "that'll cost more" or "you don't need that yet". Skills get a project built. Fit gets it built well, and keeps it that way.

When you're ready to have that conversation, the best test is a real one: tell us about your project and see how we respond. You'll learn more from one honest scoping call than from any brochure.